DISCUSSION
PAPER
October 2009
DISCUSSION
PAPER
Unlocking the €53 Billion Savings
from Smart Meters in the EU
How increasing the adoption of dynamic tariffs
could make or break the EU’s smart grid investment
By Ahmad Faruqui, Dan Harris and Ryan Hledik
Contents
Introduction
Section 1: The Potential Impact
of Dynamic Pricing................2
Section 2: How Much Demand
Response?............................3
Section 3: Overcoming Barriers
to Adoption.........................4
Section 4: The Value of Demand
Response.............................6
Section 5: The Cost-Benefit
Ratio of Investing in Dynamic
Pricing................................8
Introduction
T
he EU is poised to make a major
investment in smart meters. Already, Italy
has achieved a smart meter penetration rate
of around 85 percent and France has a rate of
25 percent. Last October the UK government
announced its intention to mandate smart
meters for all UK households by 2020. France,
Ireland, the Netherlands, Norway and Spain
are also projected to achieve nearly 100
percent smart meter installation by 2020,
and it seems likely that many other Member
States will follow suit.
Conclusion
Appendix
The Brattle Group provides
consulting and expert testimony in economics, finance
and regulation to corporations,
law firms and governments
around the world.
We have offices in Brussels,
London and Madrid, and a
growing presence in Italy. We
also have offices in Cambridge,
Massachusetts; San Francisco,
California; and Washington, DC.
For more information please
visit www.brattle.com.
Copyright © 2009 The Brattle Group, Inc.
We estimate the cost of this smart meter
investment in the EU to be €51 billion. This
investment is likely to yield improvements
in the way that electricity flows through
the grid by eliminating meter reading costs,
allowing for faster detection of power outages,
permitting remote connect/disconnect of
service and minimizing power theft. We
estimate these improvements to be worth
between €26 to 41 billion, leaving a gap of
€10 to 25 billion between benefits and costs.
In this paper, we argue that smart meters
can provide additional benefits because they
enable the provision of dynamic pricing to
customers. Examples of dynamic pricing
include real-time pricing (RTP) and criticalpeak pricing (CPP), a form of time-of-use
(TOU) pricing in which prices during the
top 100 hours of the year rise to reflect the
full cost of building and operating seldomutilized peaking capacity.1
Relying on experience from around the
globe, we estimate that if policy-makers can
overcome barriers to consumers adopting
dynamic tariffs, these benefits could be as
high as €67 billion. But if the adoption of
dynamic pricing is similar to levels seen for
other TOU tariffs in liberalised markets, then
the benefits may be only €14 billion. The
difference in benefits between high and low
adoption rates is €53 billion.
While policy-makers have to date focused
on rolling out smart meters, the issue of
ensuring that suppliers offer and customers
accept smart tariffs has received relatively
little attention. And yet overcoming barriers
to the adoption of dynamic tariffs could be
worth as much as €53 billion – sufficient to
pay back the cost of smart meter investment
for the EU, even ignoring the operational
benefits we cite above. Introducing “dynamic
by default” transmission and distribution
tariffs, stressing the environmental benefits
of dynamic tariffs and making the financial
rewards transparent to customers will be the
key to landing this €53 billion prize.
The views expressed in this paper are strictly those of the authors and do not necessarily state or reflect the views
of The Brattle Group, Inc. or its clients.
October 2009
DISCUSSION
PAPER
Unlocking the €53 Billion Savings from Smart Meters in the EU
Section 1 The Potential Impact of Dynamic Pricing
The demand for electricity is highly concentrated
in the top one percent of the hours of the year.
If a way can be found to shave off some of this
peak demand, it would eliminate the need to
install generation capacity that is used less than a
hundred hours a year. Moreover, since this capacity
is rarely used it tends to be from relatively old,
polluting plants.
For example, in most parts of the EU, between five
to eight percent of installed capacity is only used
about one percent of the time.2 In Great Britain
(GB), 6.6 percent of capacity is used one percent
of the time, and in Italy and Spain the equivalent
figures are 5.3 and 6.8 percent respectively. Figure 1
illustrates this, using load data from France for 2008.
Since electricity cannot be stored and has to be
consumed instantly, and since generation plants
of varying efficiency are used to meet demand, the
cost of power varies by time-of-day and day-ofyear. The most opportune way to try and reduce
the use of peaking capacity is to provide accurate
price signals to customers that convey the true cost
of power — that is, implement dynamic pricing.3
Once clear price signals are conveyed to customers,
they can decide whether to continue buying power
at higher prices or curtail their usage during
peak hours. This market-driven concept can save
consumers substantial amounts of money.
How much will be saved through dynamic pricing
will depend on how much peak load can be
reduced by customers, i.e., on the magnitude of
the induced demand response (DR), and how much
generation investment and fuel would be offset by
this demand response. The first item depends on
two things: how rapidly suppliers offer tariffs that
provide dynamic price signals to customers, and
how well customers respond to the price signals.
Figure 1 Top 15% of the French load duration curve, 2008
Page 2
October 2009
Unlocking the €53 Billion Savings from Smart Meters in the EU
DISCUSSION
PAPER
Table 1 Summary of the results of the main demand response trials
completed to date6
Continent and Trial
Demand Response measured
United States & Canada
Hydro Ottawa, Ontario, Canada
California-Anaheim
California-Automated Demand Response Pilot System (ADRS)
California-State-wide Pricing Pilot
Colorado-Xcel Energy TOU Pilot
Florida-The Gulf Power Select Program
Idaho Power Company Energy Watch Pilot
Illinois-Energy Smart Pricing Plan
Missouri-AmerenUE Critical Peak Pricing Pilot
New Jersey-GPU Pilot
New Jersey-PSE&G Residential Pilot Program
Washington (Seattle)-Puget Sound Energy (PSE) TOU Program
Washington-The Olympic Peninsula Project
Ranged from 5.7% to TOU-only participants to 25.4% for CPP participants
12% reduction during peak hours
Reductions as high as 51% on CPP event days and 32% on non-event days
CPP-Fixed: On critical days, average reductions at peak were 13.1%; ranged from 7.6
to 15.8%
TOU: Peak period energy use reduced 5.9%
CPP-Variable: Reduction range of 16 to 27%
TOU: Range of 5.19 to 10.63% during peak hours
CPP: Range of 31.91 to 44.81% during critical peak hours
CTOU: Range of 15.12 to54.22% during critical peak hours
41% Reduction during critical peak period
Average reduction of 5.03 kW for 4 hour event
2005: 15% reductions
2006: own-price elasticity of -0.067; -0.098 with AC
TOU: No statistically significant impact
TOU-CPP: 12% reduction in 2004, 13% reduction in 2005
TOU-CPP-Tech: 35% reduction in 2004, 24% reduction in 2005
Range from 26 to 50%
TOU: Range of 6 to 21%
CPP: Range of 14 to 26%
5% per month
15 to 17% reduction for RTP group, 20% reduction for TOU/CPP group
Australia
Country Energy, Australia
Energy Australia
New South Wales/Australia - Energy Australia's Network Tariff Reform
Reduction of 30% across peak periods
Reductions on days with a CPP event of between 5.5 and 7.8%
24% reduction for DPP high rates, 20% reduction for DPP medium rates
Europe
Norway
France-Electricite de France (EDF) Tempo Program
8 to 9% reduction at peak
Own price elasticity of -0.18 in off-peak usage to -0.79 during peak
Section 2 How Much Demand Response?
A prerequisite to the provision of dynamic pricing
is the installation of smart meters or, to use the
technical term, advanced metering infrastructure
(AMI).4 To date, AMI installation in the EU has
been relatively limited. Italy has by far the highest
installation of smart meters, with a penetration rate
of around 85 percent. The next highest is France
with 25 percent, but the majority of Member States
have penetration rates of only a few percent.
This picture will be transformed over the next ten
years. For example, the UK government announced
in October 2008 its intention to mandate smart
meters for all households by 2020. France, Ireland,
the Netherlands, Norway and Spain are also
projected to have close to 100 percent smart meter
installation by 2020, and it seems likely that many
other Member States will follow suit.
Even countries that have installed AMI systems
do not yet have dynamic pricing designs in place.
There is also considerable uncertainty as to how
customers will respond to such pricing signals.
Moreover, some policy-makers are afraid of a
customer backlash to potentially volatile prices.5
A number of trials have attempted to measure DR
via pilot programmes – we summarise the main
results in the table above. Several of these trials
use relatively simple time-of-use tariffs rather
than dynamic pricing. Nevertheless, the trials
demonstrate that, on average, customers will
respond to higher prices by lowering usage during
peak hours and by so doing, will reduce their
annual power bills.
Page 3
October 2009
DISCUSSION
PAPER
Unlocking the €53 Billion Savings from Smart Meters in the EU
Section 3 Overcoming barriers to adoption
Clearly the degree of response varies not only
by jurisdiction but by the methods used to
stimulate demand response. These range from
TOU tariffs with reminders to consumers and the
use of “traffic lights” or SMS messages to indicate
periods of high prices, to full dynamic pricing with
enabling technologies such as smart thermostats
and always-on gateway systems.
Smart thermostats automatically raise the
temperature setting on an air conditioning unit
by two or four degrees when the price becomes
critical. Always-on gateway systems turn off
appliances (e.g., washing machines and freezers)
at periods of high prices, and represent state of the
art enabling DR technology. Clearly the difference
between tariffs at different times of day also has a
strong effect on the degree of demand response.
Perhaps unsurprisingly, these studies have found
a stronger DR when more sophisticated, and more
expensive, enabling technologies are used than
when the customer still has to intervene to reduce
demand. In the state of California experiment,
the average customer reduced demand during the
top 60 summer hours by 13 percent in response to
dynamic pricing signals that were five times higher
than their standard tariff.7 Customers who had a
smart thermostat reduced their load about twice
as much, by 27 percent. Consumers who had the
gateway system reduced their load by 43 percent.8
This experiment also showed that customers did
not respond equally to the price signals. Some
responded often and others did not respond at all.
In fact, about 80 percent of the collective DR came
from just 30 percent of the customers.9
While there is now good evidence that at least
some consumers will respond to price signals, there
is still a question over how applicable these DR
experiments are to the EU in general. The studies
in the previous table has been carried out in
jurisdictions with high rates of air conditioning —
California, Ontario, Australia — or, in the case of
Norway, a high capacity of electrical space heating.
A recent U.S. study for the Federal Energy
Regulatory Commission (FERC) on the potential for
Page 4
DR in all 50 U.S. states, which The Brattle Group
coauthored, may provide some guidance as to the
potential for DR in the EU.10 The study was the
first of its kind, establishing bottom-up, statelevel estimates of both the existing amount of DR
and the potential peak reductions that could be
achieved under three distinct scenarios. A key input
to this study was each state’s saturation of central
air-conditioning (CAC). The study found that even
New England states with low CAC saturations, in
the 10 to 20 percent range, still had the potential
to reduce peak demand by up to roughly 15 percent
through cost-effective DR measures.
While the FERC study provides useful guidance
for the EU, it is still highly desirable for Member
States to undertake their own DR trials, so that
the potential benefits can be properly assessed
and weighed against the costs. For example,
Ireland has recently begun Customer Behaviour
Trials, which will involve time of use tariffs, inhome display units giving detailed information on
current prices and consumption and bills designed
to facilitate demand response and energy saving.11
Trial participants have already been recruited, and
the trials will start in January 2010.
In the GB market, Ofgem, the energy regulator,
is managing DR trials that are jointly funded by
the government and major energy suppliers. The
trials will investigate alternative ways of making
customers more aware of their energy consumption.
They will include different combinations of realtime display devices, which show energy use
in pounds and pence, smart electricity and gas
meters, additional billing information and customer
education programmes.
The trials are made up of different combinations
of these actions and are exploring the responses
of around 50,000 different households. There will
be smart meters in around 18,000 houses and realtime display devices in about 8,000 homes. Final
reporting will be complete in autumn 2010.12
France and Germany are also undertaking largescale trials, with Austria, the Czech Republic and
Spain piloting smaller smart-meter schemes.13
October 2009
Unlocking the €53 Billion Savings from Smart Meters in the EU
The rate of adoption of dynamic pricing is a
critical point. Adoption could be much lower if
customers have to actively switch to the dynamic
tariff, rather than having one as the default.
This highlights another important difference
between EU and U.S. energy markets. In most of
the U.S. there is no retail competition; rates for
households and small commercial utilities are set
by the regulator, who is free to mandate a dynamic
pricing tariff as a default. In the EU customers
would have to actively choose a dynamic tariff.
The limited literature on the topic suggests that
this difference is important. Studies show that
about 80 percent of customers would stay on
dynamic pricing if it was offered as the default rate
and that a substantially smaller number, perhaps
20 percent, would opt in on a voluntary basis.14
The UK provides another reference point. For many
years GB suppliers have offered a simple time-of-use
tariff called “Economy 7”. Electricity is substantially
cheaper during the night (defined as 01:00 to 08:00),
and customers take advantage of the lower tariff
by using night storage heaters and water heaters
on a timer. About 3.3 million of the UK’s 22 million
domestic meters households have opted for an
Economy 7 tariff.15 This is equivalent to 15 percent,
roughly consistent with the research cited above.
The U.S. provides a sobering lesson for EU
policy-makers. In Texas, which has full retail
competition and is the market that most
resembles the EU, adoption rates for dynamic
tariffs have been among the lowest in the U.S.
DISCUSSION
PAPER
The large potential difference in adoption rates
for dynamic prices begs an important question —
what can be done to reduce barriers to adopting
dynamic tariffs? The recent FERC study identified
24 potential barriers to demand response, which
were grouped into four categories:
t
Regulatory — Regulatory barriers are caused by
a particular regulatory regime, market design,
market rule or the DR programmes themselves.
t
Technical — Potential technological barriers to
implementation of DR include the need for new
types of metering equipment, metering standards
or communications technology.
t
Economic — Economic barriers refer to situations
where the financial incentive for utilities or
aggregators to offer DR programmes, and for
customers to pursue these programmes, is limited.
t
Other — These are generally related to customer
perceptions of DR programmes and a willingness
to enroll.
Arguably, several Member States are making
excellent progress in tackling the regulatory and
technical barriers. The potential stumbling blocks
that may require greater focus in the future are
the economic barriers and persuading customers
to switch to more dynamic tariffs. For some
customers, DR programmes may not provide a
sufficient financial incentive to participate, and
customers may find it hard to estimate the benefits
of switching to a dynamic tariff.
Page 5
October 2009
DISCUSSION
PAPER
Unlocking the €53 Billion Savings from Smart Meters in the EU
Customers have had similar difficulties in estimating
the benefits of energy efficiency measures.16
Customer may be risk-averse, worrying that their
bills will increase if they switch to a dynamic tariff,
rather than focusing on the potential savings.
Customers may feel that they do not know how to
shift demand to make the most of dynamic pricing,
and there may also be inertia that contributes to
low participation rates in voluntary programmes.
We have studied many successful policies to
encourage the adoption of dynamic tariffs.
For example, quantifying and stressing the
environmental benefits of dynamic tariffs can be
appealing to some customers groups. In systems
with a large capacity of wind generation, dynamic
tariffs could help shift demand to times when wind
generation is high, lowering average emissions.
Evidence in the U.S. shows that some customers
are willing to adopt dynamic tariffs based on
environmental benefits.17
Ensuring transparent and adequate financial
rewards can also help overcome customer inertia.
Where technology allows, suppliers can simplify
the choice for customers by offering to take over
their demand response for them. An example of
this is critical-peak pricing (CPP), where customers
agree to load curtailment at times of peak demand
in return for a lower flat-rate tariff. The advantage
of these schemes is that the gains for the customer
are predictable and easy to understand. Innovative
use of information technology could also help —
once AMI is in place, customers could simply enter
their meter number into a website to generate an
estimate of savings, assuming various levels of DR.
Many Member States also require suppliers to offer
a retail tariff that the regulator has approved or
set, though the idea is to phase such tariffs out
over time as retail competition matures. Setting
a dynamic regulated tariff could be one way of
increasing adoption rates. Member States that do
not have regulated tariffs could introduce them.
But this proposal illustrates the tension between
progressing with market liberalisation, and ensuring
a high adoption rate of dynamic tariffs through
a more interventionist approach to retail tariffs.
Page 6
While policy-makers cannot force customers to
adopt dynamic tariffs for their electricity, they could
mandate dynamic transmission and distribution
(T&D) tariffs, whereby the T&D charges could vary
according to when the customer uses power, and
oblige the supplier to pass these costs through
directly to the customer. The T&D charges for many
large customers already depend on when they use
power. However, smart meters would allow similar
pricing for millions of domestic customers.
Since T&D charges make up around 20 to 30
percent of household customers’ bills, a dynamic
T&D charge could by itself elicit valuable demand
response. Making part of the bill “dynamic by
default” could also encourage switching to dynamic
tariffs for the electricity commodity itself — if the
customer is already making some demand changes
as a response to dynamic T&D charges, switching to
a “full” dynamic tariff could amplify the benefits.
Section 4 the value of DR
What is the value of a reduction in demand
during critical periods? We identify several types
of benefits. First and foremost is the reduction in
the need to install peaking generation capacity.
This is a long run benefit and consists of the sum
of avoided capacity and energy costs. It can be
readily estimated based on the capacity cost of
a combustion turbine. The second benefit is the
avoided energy costs that are associated with
the reduced peak load. Third is the reduction in
transmission and distribution capacity. This is also
a long run benefit, but is harder to quantify and is
heavily dependent on system configurations that
vary regionally.
We have estimated these savings under high and
low adoption rate scenarios. Our high adoption
rate is based on the uptake seen where dynamic
pricing is the default, but it can equally be
interpreted as a scenario where policy-makers
have successfully overcome the barriers to
adopting dynamic tariffs. The low adoption rate
scenario can be interpreted as where policymakers fail to sufficiently address the barriers.
October 2009
Unlocking the €53 Billion Savings from Smart Meters in the EU
DISCUSSION
PAPER
The total value of avoided capacity costs is €4.7
billion per year for the high-case scenario, but
only €1.0 billion for the low case scenario.
As we note above, there is considerable uncertainty
in the potential for demand response per customer
in the EU. In our estimate, based on the FERC study
we assume that the average residential customer
on a dynamic pricing tariff in the EU might reduce
demand by eight to ten percent in the absence of
any enabling technology.18
When equipped with enabling technologies,
incremental increases in peak impacts can range
from 60 to 90 percent of the reduction without
technology, depending on the customer class and
technology under consideration. Note that other
studies have shown the difference for DR potential
between different classes of customers, for example
industrial users and households.
Given the uncertainties involved, in our estimate
we simply take an average DR number which
encompasses all classes of user. Based on these
findings, and under the assumption that there
is a moderate market penetration of enabling
technologies for the aggregate level of DR, it is not
unreasonable to assume an overall peak demand
reduction of ten percent for the high adoption rate
scenario and two percent for the low scenario.
To quantify the avoided capacity cost, we first
quantify the amount of capacity that will be
avoided by a reduction in peak demand and then
value it. The amount of peaking capacity that is
needed to meet this peak demand can be computed
by allowing for a reserve margin of 15 percent and
line losses of eight percent.19 We use a value of
the avoided cost of capacity of €87/kW-year.20
Using the relationship that was observed between
annual capacity and energy benefits in a recent
U.S. analysis of DR, the annual value of avoided
energy costs is estimated at €589 million and €118
million for the high and low case respectively.21
In addition, there would be a reduction in
transmission and distribution capacity needs.
As noted earlier, they are system-dependent and
difficult to estimate. However, they are unlikely
to be zero. A conservative estimate puts them at
10 percent of the savings in generation capacity
and energy costs.22 Using this estimate results in
a reduction in transmission and distribution costs
of €536 million and €107 million per year for the
high and low case respectively.
Adding up these three components yields long
run benefits of demand response of €6 billion per
year for the high-case scenario, and €1.2 billion
per year for the low-case scenario. Over a 20-year
time horizon, these savings represent a discounted
present value of €67 billion for the high-case
scenario, but less than €14 billion for the low
case scenario — the appendix shows the details
of these assumptions. In other words, efforts to
reduce barriers to the adoption of dynamic pricing
could deliver increased present value benefits of
around €53 billion.
Demand response and AMI is likely to have other
benefits as well. DR could play an important role
in increasing energy efficiency and delivering
on the EU’s commitments to reduced energy
consumption. The reduction of less efficient
peaking plants would reduce emissions during
periods of high demand. Security of supply would
improve, as DR delivers improved system reliability
and fewer blackouts and brownouts. Customers’
increased price responsiveness would make it less
profitable to exercise market power, increasing the
competitiveness of electricity markets. AMI could
significantly enhance levels of customer service.
In this assessment, we have not quantified any of
these benefits.23
Page 7
October 2009
DISCUSSION
PAPER
Unlocking the €53 Billion Savings from Smart Meters in the EU
Figure 2 Annual long run benefits of demand desponse
Section 5 The Cost-benefit ratio of investing in
dynamic pricing
How do the quantified long-term benefits compare
to the cost of installing AMI, a precondition for
dynamic pricing? Depending on the features
and the communication technology used, AMI
investment costs can range from around €70 per
meter for household customers to €450 per meter
for more sophisticated industrial meters.24 A
large portion of the cost of AMI can be recovered
through operational benefits, such as savings
in meter reader costs, faster outage detection,
improved customer service, better management of
customer connects and disconnects and improved
distribution management. We do not include these
benefits in Figure 2.
Page 8
Within the EU, Italy has the most experience with
the costs and benefits of a large scale role-out of
smart meters. Enel, Italy’s largest power company,
has installed over 30 million smart meters at an
average cost of €70 each. While the total cost of
the project was €2.1 billion, Enel estimates annual
savings of €500 million, implying that the total
cost of the project would be recovered in about five
years. Enel estimates the savings include:
t
t
t
t
A 70% reduction in purchasing and logistic costs
A 90% reduction in field operation costs
A 20% reduction in customer service costs
An 80% reduction in the costs of revenue losses
such as thefts and failures.25
October 2009
Unlocking the €53 Billion Savings from Smart Meters in the EU
In a recent study, the UK Department of Energy
and Climate Change (DECC) estimated that net
benefits for installing both gas and electric smart
meters in the domestic sector ranged from £2.28 to
£3.59 billion,26 and savings of about £1.75 to £1.76
billion for small and medium businesses.27
billion are almost three times the remaining
AMI costs to be recovered. But the case with low
adoption rates is much more marginal. Under
all but the most optimistic scenarios regarding
operational benefits, the cost of AMI may not
justify the investment.
Assuming an approximate cost of €120 per
household meter, and €450 per non-household
meter, we estimate that a total investment of €51
billion will be necessary to install AMI throughout
the EU (our calculations are in included the
appendix). This includes the “sunk” costs of meters
already installed. If 50 to 80 percent of these costs
are recovered through operational benefits, the
remaining cost of AMI is between €25 billion and
€10 billion.
Figure 3 illustrates how the benefits of demand
response can bridge the savings gap, the difference
between operational savings and the cost of AMI,
if adoption rates for dynamic tariffs are high;
the savings gap remains if adoption rates and
operational savings are low.
It seems clear that with high adoption rates
for dynamic tariffs there is a great return on
investment. The present-value benefits of €67
DISCUSSION
PAPER
In these instances, to determine the cost
effectiveness of an AMI investment it would be
necessary to further explore additional benefits
that have not been quantified in this study,
such as improved reliability, enhanced market
competitiveness and reduced rate volatility.
Figure 3 Present value costs and benefits of AMI in the EU with high
and low adoption rates for dynamic pricing
80
High adoption
rate savings from
DR
70
60
€Billion
50
Costs
Operational
savings
Low case
40
30
20
10
Savings
gap
Low adoption
rate savings from
DR
0
Page 9
October 2009
DISCUSSION
PAPER
Unlocking the €53 Billion Savings from Smart Meters in the EU
Conclusion
The adoption of dynamic tariffs could make or
break the pay-off from the EU’s investment in
smart meters. Even with operational savings from
easier meter reading and other measures at €26
to 41 billion, this still leaves a “savings gap” of
€10 to 25 billion between benefits and costs; a gap
that might not be filled if adoption is at the low
end of typical customer participation rates.
To boost adoption rate, policy-makers and energy
suppliers need to be innovative to help increase
customer participation; quantifying and stressing
the environmental benefits of dynamic tariffs,
ensuring transparent and adequate financial
rewards and offering customers a lower flat tariff in
return for providing “automatic” demand response.
For some Member States, setting a dynamic
regulated tariff could significantly increase
demand response. Countries without regulated
tariffs could implement dynamic transmission and
Page 10
distribution tariffs that would vary according to
when the customer uses power.
Based on international experience, if policymakers fail to design effective dynamic tariff
programmes, customer adoption rates for dynamic
tariffs will likely be around 20 percent. But these
rates could increase to 80 percent if policy-makers
and suppliers implement some of the policies we
highlight above.
If 80 percent of customers reduce their demand at
peak hours due to dynamic pricing, the reduction
in associated capacity and transmission costs
would be €67 billion. However, if the uptake of
dynamic tariffs is only 20 percent, then savings
are only €14 billion.
The €53 billion difference is the reward that awaits
policy-makers if they can persuade customers to
sign on to dynamic tariffs in greater numbers.
October 2009
Unlocking the €53 Billion Savings from Smart Meters in the EU
DISCUSSION
PAPER
Endnotes
1
Ahmad Faruqui, Ryan Hledik and John Tsoukalis, “The Power of Dynamic Pricing,” The Electricity Journal, April 2009.
2
The countries included are the EU 27 excluding Latvia, Lithuania, Malta and Sweden.
In addition to dynamic pricing, demand response can also be implemented by providing cash incentives to customers that encourage
them to control usage. Examples include direct load control programmes that target end uses such as central air conditioners and
water heaters, interruptible and curtailable rates that target large customers and various forms of load curtailment that are practiced
by independent system operators and regional transmission operators around the country. In this assessment, we focus exclusively on
demand response as implemented through dynamic pricing programmes. Such programmes are triggered by economic, as opposed to
system reliability, criteria.
3
AMI refers to the entire infrastructure of “smart” meters, communication networks and data management systems required for metering
information to be collected and analysed. For a further discussion and description of AMI see “Deciding on ‘Smart’ Meters: The Technology
Implications of Section 1252 of the Energy Policy Act of 2005,” Plexus Research, September 2006.
4
For a discussion of the reasons for this hesitancy, see Ahmad Faruqui, “Breaking out of the Bubble,” Public Utilities Fortnightly, March
2007.
5
For more details on these trials, see Ahmad Faruqui and Sanem Sergici, Household Response to Dynamic Pricing of electricity – a survey
of the experimental evidence, 10 January 2009. Available at: http://www.hks.harvard.edu/hepg/Papers/2009/The%20Power%20of%20
Experimentation%20_01-11-09_.pdf.
6
The 13 percent drop occurred during the six months of the summer season from May to September. Responses during the inner
summer months of June-August were a percentage point higher. The 14 percent number might be more applicable during critical-peak
conditions.
7
8
Ahmad Faruqui and Stephen George, “Quantifying Customer Response to Dynamic Pricing,” The Electricity Journal, May 2005.
Ahmad Faruqui and Sanem Sergici, Household Response to Dynamic Pricing of electricity – a survey of the experimental evidence, 10 January
2009. Available at: http://www.hks.harvard.edu/hepg/Papers/2009/The%20Power%20of%20Experimentation%20_01-11-09_.pdf.
9
“A National Assessment of Demand Response Potential,” Federal Energy Regulatory Commission Staff Report prepared by The Brattle
Group, Freeman, Sullivan & Co. and Global Energy Partners, LLC, June 2009.
10
11
“CER, Smart Metering Project Phase 1,” Information paper 2, 31 July 2009, CER/09/118.
12
See http://www.ofgem.gov.uk/Markets/RetMkts/Metrng/Smart/Pages/SmartMeter.aspx for more details.
Haney, A, Jamasb, T, Pollitt, M, “Smart Metering and Electricity Demand: Technology, Economics and International Experience,”
Electricity Policy Research Group working paper - EPRG0903, January 2009.
13
Momentum Market Intelligence, “Customer Performance Market Research: A Market Assessment of Time-Based Rates Among Residential
Customers in California,” December 2003.
14
15
Ibid. Table 6 p.14.
For example one recent study found that Irish households had one of the lowest adoption rates of energy efficient light bulbs in the EU,
despite an estimated pay-back period of just several months. See, Di Maria, Corrado, Ferreira, Susana and Lazarova, Emiliya A., “Shedding
Light on the Light Bulb Puzzle: The Role of Attitudes and Perceptions in the Adoption of Energy Efficient Light Bulbs,” 3 May 2009.
Available at SSRN, http://ssrn.com/abstract=1417948.
16
PG&E in California has enrolled roughly 75,000 customers in its Smart AC air conditioning cycling programme based on a one-time
payment of $25 and an appeal indicating that participation would be “doing one small thing” that would “actually help prevent power
interruptions and protect the environment.” (quote taken from a PG&E direct mail offer letter.)
17
On a dynamic rate with a strong price signal during the peak period, residential customers without CAC might reduce peak demand by
between eight and ten percent. A similar magnitude of impacts might be expected from medium and large C&I customers.
18
19
This includes both transmission and distribution losses.
Value used is for the Republic of Ireland. “Fixed Cost of a Best New Entrant Peaking Plant, Capacity Requirement, and Annual Capacity
Payment Sum for the Calendar Year 2009,” SEM Decision paper – SEM-08-109, 11 September 2008, p.29.
20
21
Ahmad Faruqui, Ryan Hledik, Sam Newell and Hannes Pfeifenberger “The Power of 5 Percent,” The Electricity Journal, October 2007.
This estimate is based the filing of PG&E with the CPUC on AMI. From a national perspective, we cite the US Energy Information
Administration’s estimate that transmission and distribution costs account for some 36 percent of electricity costs. See “Electricity Power
Annual,” 2007, using data from 2005.
22
For a qualitative discussion of these benefits, see our PJM-MADRI demand response study, Quantifying Demand Response Benefits in PJM,
February 2007.
23
“Capgemini Consulting and CRE – AMM for France: the complete case,” 3 October 2007. The UK Department of Energy and Climate Change
estimate the installed costs of meters for the domestic sector at between £87 and £102, and the installed cost for advanced meters for
small and medium businesses between £384 to £398.
24
25
Borghese F., “Evaluating The Leading-Edge Italian Telegestore Project,” February 2006.
“Impact assessment of a GB-wide smart meter roll out for the domestic sector,” DECC, May 2009. Benefits considered by the DECC study
include energy and peak demand reduction of 2.8 percent, avoided costs of carbon, improved infrastructure for microgeneration, avoided
costs of meter reading, reduced customer service overheads, remote disconnection and switching, reduced cost to serve customers with
pre-payment meter, and reduced theft.
26
27
“Impact Assessment of smart / advanced meters roll out to small and medium businesses,” DECC, May 2009.
Page 11
October 2009
DISCUSSION
PAPER
Unlocking the €53 Billion Savings from Smart Meters in the EU
Appendix — Estimating Demand Response Benefits
We first estimate the total costs of smart meter installation. As the costs of domestic and non-domestic
meters vary greatly, at €120 and €450 respectively, it is necessary to know the number of each required.
The number of domestic meters should approximate the number of households, which Eurostat reported
for 2007. However, we still need an estimate of the number of Small and Medium Enterprises (SMEs) which
will also require smart meters. To estimate the number of SMEs, we looked at two countries for which we
have accurate figures available: the United Kingdom and Italy. The results are presented below.
Table 2 Proportion of non-domestic meters for Italy
Value
Units
Source
[A]
[B]
[C]
No. of households in Italy 2007
No. of meters installed by Enel
Penetration rate
23,902
27,000
85%
'000
'000
Eurostat
See note
ERGEG
[D]
Implied total number of meters
31,765
'000
[B]/[C]
[E]
Implied SME meters
7,863
'000
[D]-[A]
[F]
SME meters as % of households
33%
[E]/[A]
Note: [B]: Borghese F, "Evaluating The Leading-Edge Italian Telegestore Project," February 2006.
Table 3 Proportion of non-domestic meters for the UK
Values
Units
Source
[A]
No. of households in the UK
26,649
'000s
Eurostat
[B]
Domestic meters
21,956
'000s
See note
[C]
Total meters
28,000
'000s
See note
[D]
Implied non-domestic meters
6,044
'000s
[B]-[C]
[E]
SME as % of domestic meters
28%
[D]/[B]
Note: [B],[C]: Haney, A., Jamasb, T., Pollitt, M., Smart Metering and Electricity Demand: Technology,
Economics and International Experience, January 2009, p.15.
From the above results we use 30 percent as a reasonable estimate of the number of SME meters as a
percentage of domestic meters. We also note that the number of domestic meters for the UK case differs
from the total number of households. While the reason for this is not clear, to calculate costs we use the
higher number which is more conservative.
For the 23 countries considered, the above assumptions lead to a total cost of smart meter installation of
€51 billion. The calculation is presented in Table 4.
Page 12
October 2009
DISCUSSION
PAPER
Unlocking the €53 Billion Savings from Smart Meters in the EU
Table 4 Estimate of the cost of AMI for the EU
Cost of household smart meter, € [A] 120
Cost of industrial smart meter, € [B] 450
Non-domestic meters as a % of
[C] 30%
households
Estimated total
Number of private
number of SME
households 2007
meters
Belgium
Bulgaria
Czech Republic
Denmark
Germany
Estonia
Greece
Spain
France
Italy
Luxembourg
Hungary
Netherlands
Austria
Poland
Portugal
Romania
Slovenia
Slovakia
Finland
United Kingdom
Croatia
Ireland*
Cost
household
meters
Cost SME
meters
Total cost
'000s
[D]
Eurostat
'000s
[E]
[D]x[C]
€ mln
[F]
[A]x[D]/1000
€ mln
[G]
[B]x[E]/1000
€ mln
[H]
[F]+[G]
4,439
2,866
4,219
2,365
39,291
544
4,221
16,226
26,734
23,902
187
3,810
7,202
3,536
12,933
3,852
7,381
745
1,697
2,434
26,649
1,590
1,497
1,332
860
1,266
710
11,787
163
1,266
4,868
8,020
7,171
56
1,143
2,161
1,061
3,880
1,156
2,214
224
509
730
7,995
477
449
533
344
506
284
4,715
65
507
1,947
3,208
2,868
22
457
864
424
1,552
462
886
89
204
292
3,198
191
180
599
387
570
319
5,304
73
570
2,191
3,609
3,227
25
514
972
477
1,746
520
996
101
229
329
3,598
215
202
1,132
731
1,076
603
10,019
139
1,076
4,138
6,817
6,095
48
972
1,837
902
3,298
982
1,882
190
433
621
6,795
405
382
23,798
26,773
50,571
Total
Note: *Ireland private household data from Central Statistics Office Ireland for 2006.
When calculating benefits, we considered two cases, a high uptake case and a low uptake case. The only
difference between the two is that in the high uptake case peak demand reduction is assumed to be ten
percent, and in the low uptake case it is taken to be two percent. The calculations are presented below.
Page 13
October 2009
DISCUSSION
PAPER
Unlocking the €53 Billion Savings from Smart Meters in the EU
Table 5 Assumptions in calculation of present value of DR
financial benefits — high case
Assumption/Calculation
Value
Units
Source
467,140
MW
Demand for EU 27 excluding Latvia, Lithuania, Malta and
Sweden
10%
% of peak
Calculation of Market Potential
46,714
MW
[A] * [B]
[A]
2008 EU non-coincident peak demand at
transmission level
[B]
Market potential of DR
[C]
Peak demand reduction
[D]
Reserve margin
15%
% of peak
Generally accepted industry practice
[E]
Line losses
2%
% of peak
Generally accepted industry practice
[F]
System-level MW reduction
54,795
MW
[C] * (1 + [D]) * (1 + [E])
[G]
Value of capacity
87.12
€/kW-yr
BNE Fixed Cost of a Best New Entrant Peaking Plant 2009
[H]
Value of capacity
87,120
€/MW-yr
[G] * 1,000
[I]
Total avoided capacity cost
4,774
Million €/year
[F] * [H] / 1,000,000
Assumption
[J]
Peak demand growth rate
1.7%
% per year
[K]
Annual discount rate
8.0%
% per year
Assumption
[L]
Study time horizon
20
years
Assumption
[M]
PV of €1 annuity for 20 years
11.29
€
Assumption
[N]
Energy % of generation capacity cost
12%
% of NPV
2006 Brattle DR Study for MADRI/PJM
[O]
T&D % of energy and generation capacity cost
10%
% of NPV
2006 PG&E AMI Filing
[P]
PV avoided generation capacity cost
53,900
Million €
[I] * [M]
[Q]
PV avoided energy cost
6,645
Million €
[N] * [P]
[R]
PV avoided T&D capacity cost
6,055
Million €
[O] * [P]
[S]
PV of total avoided cost
66,600
Million €
[P] + [Q] + [R]
Table 6 Assumptions in calculation of present value of DR
financial benefits — low case
Assumption/Calculation
[A]
Page 14
2008 EU non-coincident peak demand at
transmission level
Value
Units
Source
467,140
MW
Demand for EU 27 excluding Latvia, Lithuania, Malta and
Sweden
[B]
Market potential of DR
2%
% of peak
Calculation of Market Potential
[C]
Peak demand reduction
9,343
MW
[A] * [B]
[D]
Reserve margin
15%
% of peak
Generally accepted industry practice
[E]
Line losses
2%
% of peak
Generally accepted industry practice
[F]
System-level MW reduction
10,959
MW
[C] * (1 + [D]) * (1 + [E])
[G]
Value of capacity
87.12
€/kW-yr
BNE Fixed Cost of a Best New Entrant Peaking Plant 2009
[H]
Value of capacity
87,120
€/MW-yr
[G] * 1,000
[I]
Total avoided capacity cost
955
Million €/year
[F] * [H] / 1,000,000
Assumption
[J]
Peak demand growth rate
1.7%
% per year
[K]
Annual discount rate
8.0%
% per year
Assumption
[L]
Study time horizon
20
years
Assumption
[M]
PV of €1 annuity for 20 years
11.29
€
Assumption
[N]
Energy % of generation capacity cost
12%
% of NPV
2006 Brattle DR Study for MADRI/PJM
[O]
T&D % of energy and generation capacity cost
10%
% of NPV
2006 PG&E AMI Filing
[P]
PV avoided generation capacity cost
10,780
Million €
[I] * [M]
[Q]
PV avoided energy cost
1,329
Million €
[N] * [P]
[R]
PV avoided T&D capacity cost
1,211
Million €
[O] * [P]
[S]
PV of total avoided cost
13,320
Million €
[P] + [Q] + [R]
October 2009
Unlocking the €53 Billion Savings from Smart Meters in the EU
DISCUSSION
PAPER
About the Authors
Dr Ahmad Faruqui
Principal
Ahmad Faruqui recently lead a state-by-state assessment of the potential for DR for
the Federal Energy Regulatory Commission, which was filed with the U.S. Congress
in June 2009. Last year, he performed a national assessment of the potential for
energy efficiency for the Electric Power Research Institute and wrote a report on
quantifying the benefits of dynamic pricing for the Edison Electric Institute.
Dr Faruqui received his Ph.D. in Economics and M.A. in Agricultural Economics from
the University of California, Davis.
Phone: +1.415.217.1000
Email: Ahmad.Faruqui@brattle.com
Mr Dan Harris
Principal
Dan Harris is a principal based in Rome, Italy and an expert in the economics
of gas and electricity markets around the world. He has been involved in client
engagements across the full spectrum of the industry and his clients include energy
regulators, competition authorities, gas and electricity network companies, gas
buyers and sellers and electricity generators.
Mr Harris received his M.Sc. in Economics from the London School of Economics.
Phone: +39.334.260.4000
Email: Dan.Harris@brattle.com
Mr Ryan Hledik
Senior Associate
Ryan Hledik specialises in matters affecting both the electric utility sector and
the natural gas industry. He has assisted electric utilities, regulators and research
organizations in the development of innovative demand response and energy
efficiency programmes, and in the analysis and implementation of policies regarding
these programmes.
Mr Hledik received his M.S. in Management Science and Engineering from Stanford
University.
Phone: +1.415.217.1000
Email: Ryan.Hledik@brattle.com
To learn more about our expertise and our consulting staff, including our
affiliated academic senior advisers, please visit www.brattle.com.
Page 15
ECONOMIC
Cambridge
San Francisco
AND
FINANCIAL EXPERTS
Washington
Brussels
www.brattle.com
London
Madrid